James Hensman
|
85810030c5
|
small efficiency changes in rbf
|
2013-03-27 15:08:25 +00:00 |
|
James Hensman
|
adfa6de1d8
|
rbf now works in a more memory friendly fashion
|
2013-03-27 15:03:46 +00:00 |
|
Nicolo Fusi
|
99e5075488
|
Merge branch 'master' into devel
|
2013-03-14 10:25:48 +00:00 |
|
Nicolo Fusi
|
26944c736b
|
linear is now by default non-ARD
|
2013-03-14 10:25:09 +00:00 |
|
Nicolo Fusi
|
6da0feb7fc
|
added fixed effect kernel
|
2013-03-13 14:24:10 +00:00 |
|
Nicolas
|
7a9b6ad113
|
The warnings are now handeled properly in the periodic kernels
|
2013-03-12 16:50:12 +00:00 |
|
Nicolas
|
a342dc2f23
|
errors handled in Mat32
|
2013-03-12 16:43:51 +00:00 |
|
Nicolas
|
c3dae7b252
|
Fixed bug in dK_dX for the quadratic kernel
|
2013-03-12 10:43:46 +00:00 |
|
Nicolas
|
3dd62c8251
|
Few bugs fixed in the documentation
|
2013-03-11 18:45:04 +00:00 |
|
Nicolas
|
0e187b6921
|
update in the rational quadratic kernel and new the tutorial on writting kernels
|
2013-03-11 18:21:29 +00:00 |
|
Nicolas
|
cff1ad5db8
|
New rational quadratic kernel
|
2013-03-11 17:48:38 +00:00 |
|
Nicolas
|
4297a37649
|
Merge branch 'master' of github.com:SheffieldML/GPy
|
2013-03-11 17:25:22 +00:00 |
|
Nicolas
|
a012cd64a5
|
Small changes
|
2013-03-11 17:25:08 +00:00 |
|
James Hensman
|
5b86fce1b3
|
changes tie_param to tie_params
|
2013-03-11 16:46:47 +00:00 |
|
James Hensman
|
d512e9a160
|
temporarily removed a test (linear X bias)
|
2013-03-11 14:40:02 +00:00 |
|
James Hensman
|
9b11424f1f
|
implemented psi2 'cross terms' for rbfXbias
|
2013-03-11 14:28:20 +00:00 |
|
Nicolo Fusi
|
05ca5cfe6d
|
working on psi cross terms
|
2013-03-11 14:03:23 +00:00 |
|
Alan Saul
|
a866760162
|
Removed unused partial1
|
2013-03-11 12:39:44 +00:00 |
|
Alan Saul
|
95cedc7e4e
|
fixed merge conflicts
|
2013-03-11 12:26:44 +00:00 |
|
Alan Saul
|
f562d6cd46
|
deprecated flapack, namespace changed to lapack.flapack
|
2013-03-11 12:21:12 +00:00 |
|
Alan Saul
|
12d6f5056b
|
removed keyname partial
|
2013-03-11 12:15:59 +00:00 |
|
James Hensman
|
e75b566f1e
|
Merge branch 'master' of github.com:SheffieldML/GPy
|
2013-03-11 10:43:23 +00:00 |
|
James Hensman
|
393662b05d
|
sometidying of the psi statistic cross terms
|
2013-03-11 10:43:17 +00:00 |
|
Nicolas
|
ec748e2d6b
|
all the product_orthogonal have been changed to prod_orthogonal for consistency
|
2013-03-11 10:33:29 +00:00 |
|
James Hensman
|
f881e65761
|
coregionalisation seems to be a go-go
|
2013-03-08 18:21:29 +00:00 |
|
James Hensman
|
af510d166a
|
some changes to product_orthogonal
dKdiag_dX is now implemented, some of the cod eis a little tidier
|
2013-03-07 16:01:00 +00:00 |
|
James Hensman
|
9a97ad7348
|
debugging the coregionalisation kern
|
2013-03-07 13:49:54 +00:00 |
|
James Hensman
|
d00d10952c
|
more messing around with coregionalize
|
2013-03-06 16:25:29 +00:00 |
|
James Hensman
|
65f9c7bb76
|
First attempt at making coregionalise work with the sparse model
Gradients are failing! have implemented prod_othogonal.dKdiag_dtheta
|
2013-03-06 15:29:03 +00:00 |
|
James Hensman
|
613aae6417
|
coregionalisation
|
2013-03-06 13:15:15 +00:00 |
|
James Hensman
|
f2ce47d96e
|
Added symmtrical covariance functions
|
2013-03-05 15:58:03 +00:00 |
|
James Hensman
|
c5f4bcd02b
|
added dKdiag_dtheta for linear
|
2013-03-01 23:06:58 +00:00 |
|
Nicolas
|
3ffa8d782c
|
indentation for dKdiag_dtheta fixed
|
2013-03-01 17:45:41 +00:00 |
|
Nicolas
|
adfe587c86
|
added dKdiag_dtheta for the periodic kernels
|
2013-03-01 16:52:37 +00:00 |
|
James Hensman
|
4d79c3c97d
|
start of psi2 crossterms
|
2013-02-26 14:49:00 +00:00 |
|
Nicolas
|
5f63ad5524
|
Sparse-GPLVM now seems to work beautifully with product kernels
|
2013-02-21 18:41:59 +00:00 |
|
James Hensman
|
a166187231
|
Merge branch 'master' of github.com:SheffieldML/GPy
|
2013-02-21 14:27:08 +00:00 |
|
James Hensman
|
304f9b50a8
|
fixed the bug where we couldn't tie parameters at the kern level
|
2013-02-21 14:26:36 +00:00 |
|
Nicolas
|
37e7e11b5b
|
The product by default is now on the same space
|
2013-02-20 17:50:33 +00:00 |
|
Nicolas
|
a0eca5ec9a
|
Added the product of kernels defined on the same space + a few bugs fixed in the prod_orthogonal
|
2013-02-20 17:44:40 +00:00 |
|
Nicolas
|
1ccf061b8f
|
New features in the product_orthogonal of kernels
|
2013-02-20 16:38:23 +00:00 |
|
Nicolas
|
a4939adcb0
|
the 2D plot can now handle *args, **kwargs
|
2013-02-20 14:18:45 +00:00 |
|
Nicolas
|
421b87a07c
|
small changes to the lengthscales such that the periodic kernels can be called as the non-periodic ones
|
2013-02-20 14:12:29 +00:00 |
|
Nicolas
|
336b15cb79
|
reinsert the plot function for kernel that diserpered at one point...
|
2013-02-18 10:38:29 +00:00 |
|
Nicolo Fusi
|
de665361a7
|
all kernels working fine with the psi statistics now
|
2013-02-15 18:08:40 +00:00 |
|
Nicolo Fusi
|
bf2d641112
|
debugging BGPLVM
|
2013-02-15 16:31:51 +00:00 |
|
Nicolo Fusi
|
2abba8cf14
|
merged with master
|
2013-02-15 11:56:37 +00:00 |
|
Nicolo Fusi
|
8761f72c59
|
merged
|
2013-02-12 18:04:07 +00:00 |
|
Nicolo Fusi
|
03c1f77c08
|
trying to fix bugs in kerns
|
2013-02-12 17:45:44 +00:00 |
|
Nicolò Fusi
|
baa06181e8
|
psi statistics working for linear ARD kernel
|
2013-02-07 10:01:05 +00:00 |
|